Background
Rheumatoid Arthritis (RA) is a chronic autoimmune disease, with a worldwide adult prevalence of 0.2–1.2 % [
1]. The disease is painful and progressive, leading to increasing levels of disability and systemic complications [
2]. There is currently no cure for RA: treatment aims are to reduce pain and inflammation, delay joint erosion and maintain function [
3]. Depression and anxiety are highly prevalent in rheumatoid arthritis (RA). According to a recent meta-analysis [
4], 14.8 % of patients screen positive for depression when measured with the Hospital Anxiety and Depression Scale (HADS [
5]) with a threshold of 11 or more indicating the presence of depression. This prevalence estimate is substantially higher than the 5.0 % level reported in the general population [
6]. Symptoms of anxiety are also frequently reported in RA, with 25.1 % of RA outpatients screening positive for anxiety according to the 7-item Generalised Anxiety Disorder (GAD-7 [
7]) questionnaire [
8]. Anxiety and depression are sometimes grouped together and referred to as “common mental disorders”. Anxiety and depression may also be seen as symptoms which are present with varying degrees of severity in the population. In this paper, we refer to depression and anxiety scores as continuous indicators of symptom severity, rather than as disorders reaching a threshold of severity. This method is particularly beneficial as it maximises statistical power, and allows examination of the linear associations between mental health and disease outcomes [
9,
10].
RA has significant implications for patient quality-of-life [
11] and increased psychological symptoms in RA is associated with poorer patient outcomes including increased pain [
12], fatigue [
13], service usage [
14] and increased risk of premature mortality [
15]. The relationship between mental and physical health is bidirectional. Experiencing psychological distress may inflate the subjective severity of patient-reported symptoms such as pain and tenderness [
16]. Additionally, psychological distress may impact health outcomes by influencing health behaviours such as medication adherence [
17] and smoking [
18]. Reduced levels of physical activity can result in de-conditioning, loss of natural endorphins and increased pain [
19]. Furthermore, common mental disorders are associated with immune dysregulation [
20‐
22].
A recent systematic review has found that only seven studies have examined the longitudinal relationships between depression and RA outcomes [
23]. This review concluded that depression may worsen pain and disease activity, and reduce treatment efficacy, although evidence is limited by a small number of studies, often of poor quality. Studies frequently lacked a priori hypotheses, used convenience samples, and had inadequate adjustment for confounders [
23]. We previously reported that increased psychological distress predicts increased disease activity (measured via the 28-joint Disease Activity Score; DAS28 [
24]) and reduced odds of reaching clinical remission over a 2-year follow-up period, however this research is limited by its sub-standard identification of psychological distress, and its use of clinical trial data representing a relatively homogenous group of patients [
25].
More research is needed to resolve these issues and replicate previous research findings, using robust methods of identifying depression and anxiety in a naturalistic patient sample. This could have substantial implications for the treatment of RA: if depression and anxiety are drivers of the non-inflammatory components of DAS28 disease activity, optimal RA treatment plans could involve the pharmacological or psychological management of depression and anxiety alongside RA. We therefore aimed to explore the relationship between depression and anxiety and disease activity after one year. We tested three hypotheses: 1) that increased depression and anxiety symptom severity at baseline would be significantly associated with increased DAS28 at 1-year follow-up; 2) that the relationship between depression and anxiety and DAS28 would be primarily driven by an association between depression and anxiety and the subjective DAS28 components: tender joint count (TJC) and patient global assessment (PGA); and 3) that increased levels of depression and anxiety at baseline would be associated with reduced odds of reaching clinical remission at 1-year follow-up.
Discussion
We have found evidence to support one of the three hypotheses: 1) depression and anxiety scores at baseline were not significant predictors of DAS28 disease activity at 1-year follow-up after full adjustment; 2) depression and anxiety were significantly associated with the subjective components of the DAS28: TJC and PGA; 3) depression and anxiety were not significantly associated with the odds of reaching clinical remission at 1-year follow-up after adjusting for covariates. These results support our previous findings from clinical trial data showing a longitudinal relationship between baseline depression/anxiety and follow-up tender joints, patient global assessment, however we failed to replicate our previous finding of an association between depression and anxiety and disease activity [
25]. The current study uses a more robust, validated method of identifying depression and anxiety, in a more heterogeneous clinical sample, with longer and more variable disease duration. This strengthens the evidence indicating a prospective relationship between depression and anxiety and poorer subjective disease outcomes in RA [
23]. Our failure to replicate the association between depression and anxiety and odds of reaching clinical remission may be due to a lack of statistical power to predict a binary outcome.
There are several possible explanations for the associations found between depression and anxiety and patient global assessment and tender joint counts. The negative cognitions often experienced in depression and anxiety [
31] may contribute to how RA patients interpret and perceive their symptoms [
32]. Psychological distress is associated with worsened health behaviours such as failure to take medications as prescribed [
17], and increased smoking [
18], or may contribute to reduced physical activity [
19], which can also contribute to worsened disease outcomes. The association between depression and anxiety and the subjective DAS28 elements would suggest that depression/anxiety impact perceptions and behaviours, rather than immune dysregulation. Further examination of mediators in this relationship is warranted and potential targets for investigation are negative cognitions, behavioural activity and health behaviours.
Increased levels of tenderness and poor patient global assessments, despite well-managed inflammation, can inflate DAS28 scores thereby reducing the perceived efficacy of treatment [
28]. There is substantial evidence to suggest that depression and anxiety can be effectively treated in physical conditions [
33,
34]. Given the importance of the DAS28 in clinical decision making, the routine detection and management of depression and anxiety may be a further strategy to improve disease management [
8], and doing so would align with National Institute of Health and Clinical Excellence guidelines [
35].
Limitations
This study has some limitations to consider. Firstly, the sample size is small, limiting statistical power and the scope to control for pertinent covariates. Our selectivity about covariates, whilst evidence-based, may have influenced results, and inclusion of several other variables such as comorbidities and ethnicity would have been preferable. Future research would benefit from recruiting patients with a clinician-verified RA diagnosis, rather than relying on self-reported diagnosis. We have shown that only 51.4 % of patients with a self-reported RA diagnosis in fact, have clinically verified RA. A further consideration was the lack of sociodemographic data available; low socioeconomic status (SES) patients are typically under-represented in research samples [
36]. As low SES is associated with increased susceptibility to depression [
37] and RA [
38], our findings may not be generalizable to the general RA population, although it is important to note that KCH caters mostly for patients in South East London, which has a higher level of deprivation than the England average [
39,
40]. Additionally, we were unable to invite patients who could not read/write English to complete the questionnaires, which may further limit the generalisability of our findings to the wider, non-English-speaking population.
Additionally, no information was collected during recruitment about patients who declined to participate. Therefore we were unable to determine participation rate, examine any demographic, physical, or psychological determinants of non-participation, or examine between-group differences in patients who consented to participate and those who did not. Future replication of this study should attempt to record details for all patients approached for recruitment.
A final limitation is the lack of data available regarding medication usage throughout the follow-up period, for either RA or mental disorder. DAS28 outcomes may be substantially driven by treatment intensity and modality, and treatment decision-making may be influenced by patient mental-state [
41]. The addition of medication data may add valuable information to the understanding of these relationships. Furthermore, the inclusion of DAS28 data at only one follow-up point means that any variation in disease activity throughout the follow-up period, or as a result of treatment, cannot be assessed.
Future research may also benefit from taking into account change in depression and anxiety. There is some prospective evidence suggesting initial depression levels predict between 37 and 58 % of the variance in follow-up mood scores in RA [
42]. With a larger sample size, it would be interesting to stratify our results by change in mood over time, to see if the relationship between mood and disease activity alters by mood trajectory.
Conclusions
This study supports the findings from previous research indicating an association between depression and anxiety scores at baseline and worsened disease outcomes [
23,
25]. These findings have several implications. Inflated DAS28 in the context of clinically well controlled disease may indicate significant psychological morbidity rather than true disease activity. Regardless of direction of causality, the consistent association between depression and anxiety and disease variables is strong, and depression and anxiety may act as easily identifiable and manageable “psycho-markers” of adverse disease outcome [
43]. We recommend that depression and anxiety be measured in routine clinical practice and as part of randomised controlled trials for new treatments in RA.
Finally, research is required to determine whether effective treatment of depression and anxiety can improve rheumatological outcomes. Psychological interventions have been used to successfully improve disease outcomes in diabetes and coronary heart disease [
44], and our results highlight the need to test a similar approach in rheumatoid arthritis.
Ethics and consent to participate
Participants gave informed consent, and all study procedures and the study protocol was approved by the South East London Research Ethics Committee (REC reference number: 10/H0808/135).
Consent to publish
Not applicable.
Availability of data and materials
All supporting data can be provided upon request to the authors.
Acknowledgements
We also thank Putu Khorisantono, James Gwinnutt, Fatma Mehmet and Egli Ioannou for assistance with data collection, along with Radka Chura for assistance with accessing data from patients’ hospital records.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
FM designed the analysis and developed the hypotheses, as well as performing all analyses and writing the manuscript. SA assisted with data collection and data management, as well as reviewing and commenting on all iterations of the manuscript write-up. KI assisted with the design of the study and interpretation of results and reviewed and commented on all iterations of the manuscript write-up. MH contributed to the analysis design, assisted with analysis and reviewed and commented on all iterations of the manuscript write-up. TC contributed to the acquisition of data, and development of analysis methodology and reviewed and commented on all iterations of the manuscript write-up. All authors read and approved the final manuscript.